marrow transplant models and the established use of bone marrow transplantation in human therapies against cancer and other hematopoeitic disease; and we decided to examine muscle cells as a target for antibody production due to their easy accessibility and high protein synthetic capacity. This chapter will focus results on the work done on the hematopoietic cells that are not of the B or T lineage. The next chapter will describe our work on muscle cells.
Materials and Methods pMIG-aHEL Vector
The aHEL IgG1 antibody is contructed by PCR cloning the entire κ light chain and the heavy-chain-variable region of the anti-HEL antibody from the MD4 mouse genomic DNA. The light chain is fused to the heavy-chain-variable-region DNA via a F2Aopt element (described in Chapter 2). The light chain-F2Aopt-heavy-chain-variable-region DNA is then grafted onto a murine anti-human CD34 IgG1 antibody by SOE (splicing- by-overlapping-extension) PCR. The cassette is then cloned into the pMIG vector between the Not1 and BamH1 sites.
Tranduction of Murine Bone Marrow Cells and Transplantation of the Transduced Cells Transduction of murine bone marrow cells and transplantation of the transduced cells were carried out as described by Yang et al. (Yang and Baltimore 2005). Briefly, we harvested bone marrow from donor Rag1 mice and cultured the cells in DMEM plus 10%
heat-inactivated fetal bovine serum supplemented with recombinant murin IL-3 (20 ng/ml), IL-6 (50 ng/ml), and SCF (50 ng/ml). The bone marrow cells were transduced with the pMIG-aHEL retroviral vector by spin-infection once per day for three days beginning 24 hours after the harvest. The cells were injected into irradiated Rag1
recepient mice one day after the last spin-infection. The Rag1 mice were pre-conditioned with 300 rads of radiation from a Cs137 source.
Serum Collection and Mouse IgG ELISA
75 ul of blood was collected from each mouse each time using a heparin-coated
microcapillary tube by retro-orbital bleeding and transferred into a microcentrifuge tube and kept on ice. The blood was then incubated at 37°C for 30 mins, and then spun down at 1150g at 4°C. The serum was collected from the top of the tube. Murine IgG ELISA was performed using a Mouse IgG ELISA Kit (Bethyl Laboratories Inc., Montgomery, TX) per manufacturer’s instructions.
Results and Discussion
The Use of Non-Lymphoid Hematopoietic Cells to Produce Antibodies
We asked whether it was possible to produce antibodies in non-lymphoid hematopoietic cells. We employed an in vitro and in vivo models in parallel.
Specificially, we were interested to see whether myeloid cells could be made to produce
antibodies. For the ex vivo studies, we chose to use bone-marrow-derived-macrophages (BMMs) as a representative myeloid lineage cell because the culture conditions for them are well established (Zhang, Goncalves et al. 2008). They were chosen also because they up-regulate the Xbp-1 gene on activation by LPS (Martinon, Chen et al. 2010; Zeng, Liu et al. 2010; Dickhout, Lhotak et al. 2011), a gene that is required for and similarly induced in the B-cell-to-plasma-cell transition, and required for the secretion of immunoglobulins (Reimold, Iwakoshi et al. 2001; Calfon, Zeng et al. 2002; Iwakoshi, Lee et al. 2003).
Human antibodies interact differently than mouse antibodies with murine Fc receptors. To more faithfully monitor the effects that antibody expression might have on mouse immune cell function, we chose to use a mouse antibody instead of a human one in this study. We cloned a mouse anti-HEL IgG1 gene into the retroviral vector, pMIG (Figure 4.1A). We harvested mouse bone marrow cells and cultured them in the presence of 10 ng/ml of M-CSF (macrophage-colony stimulating factor), a culture condition conducive to the differentiation and survival of macrophages from bone marrow stem and progenitor cells. During the culture period, the cells were transduced by spin-infection once per day, for four days (Figure 4.2B). At the end of the ten-day culture, we changed the media by carefully removing most of the media from each well. At this time, only adherent cells were kept. Cells were challenged with 10 ng/ml lipopolysaccharide (LPS) for two days, in the absence or presnce of M-CSF for one or two days. We observed a marked transition of macrophage morphology 24 hours after the LPS challenge, with the
cells taking on an activated, more spread out morphology and many fewer cells staying in the spherical morphology. We quantitated the amounts of antibody in a 24-hour period for two days by aspirating the media, washing the cells, and adding fresh media to the cultures. We found that the macrophages were able to produce a substantial amount of antibody (Figure 4.2). To give a sense of scale, OCI-Ly7 cells, a DLBCL (diffuse large B cell lymphoma) B cell line, when transduced with a human antibody gene by lentiviral vectors, are capable of producing on the order of 50 ng/ml of IgG in a culture volume of 10 ml containing a total of roughly 5 million cells over a period of 48 hours, giving an output of 50 ng/ml × 10 ml / 5 million cells / 2 days = 5 × 10-14 ng/cell/day. In the case of the BMMs, the number of cells at the end of the 10-day culture approximately quardruples the initial number of input bone marrow cells at the start of the culture, giving us approximately 2 million cells in each 1 ml culture in the 24-well plate. The number of cells reaches a plateau at this time due to near confluent growth. Thus, the per- cell antibody output from the macrophages can be estimated to be 50 ng/ml × 1ml / 2 million cells / 1 day = 2.5 ×10-14 ng/cell/day. This estimate of the antibody production capacity of an activated macrophage is on the same order of magnitude as that of a B- lineage cell that constitutively secretes antibodies. This is rather impressive.
This level of antibody output bodes well for the use of engineered non-lymphoid hematopoietic cells to produce antibodies, but here it also raises a legitimate concern that the high levels of antibody output from the cells that do not normally make antibodies might disrupt the normal functioning of these cells, by, e.g., impacting the synthesis and
secretion of other crucial proteins, such as secreted cytokines. Extensive in vivo characterization would be required to fully address this concern, but to get an initial handle on the issue, we quantitated the amount of the inflammatory cytokines IL-6 and TNFα secreted by the transduced BMMs on LPS-challenge as a measure of the overall function of these cells. We found that the amounts of cytokines secreted by the antibody- vector transduced cells were not significantly different from those produced by cells transduced by control-vectors carrying either IRES-GFP alone or a Luciferase-IRES-GFP cassette (Figure 4.3). Taken together, these data suggest that we might be able to use non- lymphoid cells to produce a substantial amount of antibody in vivo. To test this
hypothesis, we turned now to a bone marrow adoptive transfer model.
Figure 4.1 A) Schematic representation of the pMIG-aHEL-mIgG1 retroviral vector. B) Experimental design for the BMDC model of ectopic expression of antibody genes in the non-lymphoid hematopoietic cells. After transduction and culture for ten days, the cells were challenged with LPS in the absence or presence of continued M-CSF stimulation.
Figure 4.2 Antibody production by bone marrow-derived-macrophages transduced with pMIG-aHEL antibody vector or pMIG control vector.
Figure 4.3 Cytokine production from vector-transduced bone-marrow-derived- macrophages.
To study whether non-lymphoid hematopoietic cells could be made to produce antibody in vivo, we chose to use the Rag-1 knockout (Rag1) mouse radiation chimera model. The Rag1 mouse does not produce any functional T or B cells due to
homozygous deletion of the RAG1 gene. Therefore, antibody measured in blood would have to have been derived from the antibody vector. The strain of Rag1 mice that we propagated in the lab was more sensitive to ionizing radiation than a Rag1 strain based on C57/BL6 mice, commercially available from Jackson Laboratory (Bar Harbor, ME). The lethal dose for the strain we used was achieved at 450 rads, due to modes of radiation death from organ damage not related to bone-marrow failure; this is in contrast to wild- type C57/BL6 mice and C57/BL6-based Rag1 mice, which can tolerate a lethal dose of up to 900 rads, and die due to bone-marrow failure. We therefore chose to use a dose of 400 rads to maximize clearance of endogneous marrow while preserving the viability of the transplant recepients. This “sublethal” dose of radiation is not myeloablative in the Rag1 mice.
The experimental design is outlined in Figure 4.4A. 5-FU-enriched bone marrow cells were cultured in cytokine-enriched media and transduced with the pMIG-aHEL vector by spin-infection. The cells were injected into irradiated Rag1 hosts. The animals were bled weekly, starting four weeks post-transplant, to monitor serum IgG levels, and on week 4 and week 8, for both IgG ELISA and FACS analysis to assess the degree of engraftment (Figure 4.4B and C). We found that three animals in the experimental group that received the pMIG-aHEL treated cells had relatively high levels of reconstitution
(1130B, C, and D) as assessed by GFP positivity to be around 10–20% of the peripheral blood leukocytes. These animals maintained high levels of antibody concentrations (mean: 350 ng/ml). One animal, 1130A, had a low level of reconstitution (1.68% at week 4, and 1.49% at week 8), and had a low level of serum antibody (mean: 25 ng/ml). We observed that the levels of antibody produced appeared to be correlated with the degree of reconstitution as assessed by the percentage of cells that expressed GFP. The levels in the serum rose and fell with the levels of reconstitution.
Figure 4.4 A) An outline of experimental design for studying antibody production from non-lymphoid hematoipoietic cells in pMIG-aHEL transduced Rag1 mice. B) Serum antibody levels post-transplant. The time shown is the number of weeks after bone marrow transplant. UTS: Untransduced control mice. 1130 A–D: pMIG-aHEL
transduced-bone-marrow recipient mice. C) Percent reconstitution by vector-transduced cells.
The normal concentrations of IgG antibody present in mouse serum is
approximately between 1–10 mg/ml, depending on a variety of factors, including strain, gender, and age. Thus the levels of antibody we achieved represented less than 0.035%
of the total serum antibody concentration in the mouse. To give additional perspective, a previous attempt by Dr. Lili Yang in our laboratory to express antibodies in murine bone marrow transplant models that also employed the pMIG vector resulted in a serum concentration of approximately 1 ug/ml in wild-type mice, which had normal B and T cells. We were thus able to achieve approximately 40% of that level using non-lymphoid cells alone. We note that this value was achieved with only 10% of chimerism. We believe that the antibody concentrations could be higher if the levels of engraftment could be improved in wild-type animals that can tolerate myeloablative preconditioning prior to transplantation.
In summary, we showed that non-lymphoid hematopoietic cells are viable alternative cell types to lymphocytes for engineering antibody production. The
engineered myeloid cells (BMMs) are functionally normal. The Rag1 BM transplanted animals produced sustained and high levels of antibodies from non-lymphoid cells, and the levels of antibody production are correlated with the degree of transduced cell engraftment. The expression of antibody from engineered non-lymphoid hematopoietic cells thus represents a potentially useful alternative to engineering B cells.
References
Calfon, M., H. Zeng, et al. (2002). "IRE1 couples endoplasmic reticulum load to secretory capacity by processing the XBP-1 mRNA." Nature 415(6867): 92-96.
Dickhout, J. G., S. Lhotak, et al. (2011). "Induction of the unfolded protein response after monocyte to macrophage differentiation augments cell survival in early
atherosclerotic lesions." FASEB J 25(2): 576-589.
Iwakoshi, N. N., A. H. Lee, et al. (2003). "Plasma cell differentiation and the unfolded protein response intersect at the transcription factor XBP-1." Nat Immunol 4(4):
321-329.
Martinon, F., X. Chen, et al. (2010). "TLR activation of the transcription factor XBP1 regulates innate immune responses in macrophages." Nat Immunol 11(5): 411- 418.
Reimold, A. M., N. N. Iwakoshi, et al. (2001). "Plasma cell differentiation requires the transcription factor XBP-1." Nature 412(6844): 300-307.
Yang, L. and D. Baltimore (2005). "Long-term in vivo provision of antigen-specific T cell immunity by programming hematopoietic stem cells." Proc Natl Acad Sci U S A 102(12): 4518-4523.
Zeng, L., Y. P. Liu, et al. (2010). "XBP-1 couples endoplasmic reticulum stress to augmented IFN-beta induction via a cis-acting enhancer in macrophages." J Immunol 185(4): 2324-2330.
Zhang, X., R. Goncalves, et al. (2008). "The isolation and characterization of murine macrophages." Curr Protoc Immunol 14:11.